资源约束下基于离散粒子群算法求解多Agent资源结盟博弈问题  

Solving Coalitional Resource Games Based on Discrete Particle Swarm Optimization under Resource Constraint

作　　者：骆剑彬[1] 胡山立[1] 林耀海[1] 

机构地区：[1]福州大学计算机科学与技术系,福州350002

出　　处：《计算机与数字工程》2008年第10期12-16,共5页Computer & Digital Engineering

基　　金：国家自然科学基金项目(编号:60573076)资助

摘　　要：针对资源结盟博弈(CRGs,Coalitional Resource Games)中自治Agent结盟问题,每个Agent占有的资源不足以满足其目标要求,Agent通过建立联盟可以相互合作,各自实现目标。求解CRGs稳定联盟及其可达目标集的问题是一个NP-完全问题,由于可能的联盟数目随着Agent数目呈指数增长,需要设计高效的搜索算法。通过引入离散粒子群算法解决CRGs中自治Agent结盟问题。We study Coalitional Resource Games（CRGs）,in which each agent is endowed with a set of resources which is not enough to realize its goals. By colition,agent are collectively endowed with the necessary resources to satisfy their goals. Searching the stable coalition is an NP- complete problem. Due to the exponentially increase of the number of possible coalitions, a heuristic efficient algorithm is required. Then a particle swarms cooperative optimization algorithm applied to solving coalitional resource games under resource constraint.

关 键 词：资源结盟博弈 AGENT 离散粒子群算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]